Primary care is primed for AI. But is AI aligned with primary care?

Fewer than one-third of primary care clinicians have a say in selecting the AI products their institutions expect them to fold into their clinical workflows. That’s a problem. 

Why? Because the degree to which AI tools get appropriately deployed in real-world settings depends greatly on buy-in levels from end-users. And technology resistance can fester and spread among workers in any field who feel changes have been foisted on them from above.  

The conductors of a new survey—the nonprofit org Rock Health and the American Academy of Family Physicians—make the point in a report published Feb. 24. 

The survey went out last fall to more than 1,200 family doctors and other PCPs. Here are excerpts from the report. 

1. Family physicians and other primary care providers (PCPs) are actively exploring AI both personally and professionally. 

Half of respondents reported having used AI tools for at least one use case at work. Some 62% said they use generative AI tools like ChatGPT outside of work; among those that haven’t, 80% expressed a desire to do so.

‘We might have expected that residents and early-career physicians (those within 10 years post-residency) would lead the charge adopting AI tools. However, early-career respondents were no more likely to have tried AI solutions than their more established colleagues.’

2. Close to 1 in 5 PCPs use AI tools daily in their personal lives.

This outpaces the adoption rate of the general U.S. population, where only 6.4% report daily personal AI use.

‘Workwise, more than 90% of survey respondents have tried or are open to trying a variety of professional AI categories—including more nascent AI use cases such as behavioral health support and population health management.’

3. Most PCPs have limited influence over the AI tools available in their practices. 

Almost two-thirds, 65%, said they have limited or no input on AI purchasing decisions. The gap between AI buyers and users isn’t entirely surprising considering how the practice landscape has shifted from physician-owned and -led practices toward more of an employment model.

‘Given the high costs, inherent risks and desire for consistency across organizations, top-down decision making is the norm. However, this also creates vulnerability: AI’s long-term success in primary care will ultimately depend on the buy-in of those using the tools.’

4. PCPs are optimistic about some aspects of AI but skeptical about others. 

Many said they believe AI will improve their own clinical efficiency over the next three  years, mainly by improving time to diagnosis (73% expect positive impacts), diagnostic accuracy (66%), and appropriateness of treatment plans (66%). Most expect AI will improve their wellbeing (70%) and workload (66%).

‘Despite these expected benefits, respondents are skeptical that patients will experience the same upside. About half said they don’t expect gains from AI to positively impact patient satisfaction (47%), the personalization of care (52%) or clinician-patient relationships (55%) over the next three years.’

5. Family physicians and other PCPs are concerned about AI in the workplace. 

Eighteen percent fear that AI will have negative or very negative impacts on their job security, while another 52% are still unsure. Eighty-one percent want more training to fully trust AI solutions at work, signaling hesitancy to adopt these tools into their workflows. 

‘Nearly 70% also want medico-legal protections before they trust AI tools and 64% want education on legal, liability, and malpractice risks—likely reflecting concerns about increased liability. Meanwhile, 68% are looking for ethical guidelines on how to best use AI in their practice.’

In their concluding section, the authors note that 39% of respondents reported burnout—and most of them seem open to trying new tools for relief. 

“That said, trying is different than fully trusting, and most want additional training and legal protections before embracing AI,” the authors reiterate. More:

‘Since most physicians and other clinicians aren’t involved in AI purchasing and rollout, sustaining adoption will require concerted efforts to actively engage the clinical workforce.’

Full survey report here.